+10KW induction heater / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

+10KW induction heater

previous 1 2 3 4 next

Move Thread

LAN_403

Patrick

Mon Jul 16 2012, 08:01AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Linas wrote ...

Can i ask you for the dimensions on this ferrite core? And do you know that 30kW number with high confidence?

Linas

Mon Jul 16 2012, 09:40AM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Patrick wrote ...

Can i ask you for the dimensions on this ferrite core? And do you know that 30kW number with high confidence?

It is from CT scanner power supply for HV tube. original design had 3 cores in parallel, i got only two.
it is 12cm^2 core

i made CT part as in uzzors induction heater, and having same problem.

CCO part is simple, if E/A is high, that gives ~21kHz and if E/A is low, it gives up to 50KHz ( depend on resistor, usually i set around 50-70kHz)

i am using right now LM311 because for some reason in data sheet they say that e/a works from 1.5 to 5v, so to be sure i use LM311 since is perfect because has open collector for pull down resistor
and yes, i use E/A as comparator

just made voltage feedback from LC circuit with voltage divider, few 1N5819.
and it's works, i don't get any noise .
interesting part i can't use any capacitor on voltage feedback because it kill feedback, even as small as 330pF.
Only problem is jumping below resonance,if that part will be sorted out, that will be winner

...

Mon Jul 16 2012, 04:22PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

Like steve said, your best bet is to limit it so that it can't go below resonance, because designing a feedback network that can handle an inflection point like that will be very hard (especially when the point moves around!).

I like the suggestion of just setting a pessimistic limit and never running at 100% full power, however since the turning point moves around I can see this being problematic. Perhaps you can add a phase comparator between coil current/voltage (simple XOR gate with a low pass filter would work) and when you find that you are getting close to 0 phase difference disable the current feedback and wait for the setpoint to fall below the measured current? Perhaps turn on a 'output at maximum' light on the front panel as well

quite the project you have got going there in any case!

Linas

Mon Jul 16 2012, 05:23PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

... wrote ...

Like Steve said, your best bet is to limit it so that it can't go below resonance, because designing a feedback network that can handle an inflection point like that will be very hard (especially when the point moves around!).

it's impossible to set Fo and get power at the same time, because as you said it changing to much .
So far so good, blow 16A fuses dozens of times and IGBT are still alive :) ( and it's still 2x smaller than in final version )

I guess i have to shift to PLL driver since i get some feedback working, it's not ideal but what can i do... ?
As far as i know PLL will lock up at Fo and don't bellow, right ?

Steve Conner

Tue Jul 17 2012, 08:25AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Well there's your problem, you can't use a comparator as an error amplifier.

If you try to use them in the linear mode they just oscillate. Try again with the 33067's onboard error amp, or an op-amp like the LM324.

Linas

Tue Jul 17 2012, 08:52AM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Steve Conner wrote ...

Well there's your problem, you can't use a comparator as an error amplifier.

If you try to use them in the linear mode they just oscillate. Try again with the 33067's onboard error amp, or an op-amp like the LM324.

trust me i tried to use that, but get same results as with LM311
anyway i will try to do that now, fingers crossed.

Steve Conner

Tue Jul 17 2012, 09:35AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Well try again and pay special attention to the compensation networks shown in the MC33067 datasheet. You have to limit the gain of the error amp quite severely at high frequencies, or it'll amplify switching noise so much that it drives itself into clipping. (I guess that was happening since you got the same results with an op-amp as with a comparator.)

In this application I'd expect to see it configured as an integrator with a time constant of 10s to 100s of milliseconds: fast enough to track workpiece insertion and removal and Curie point, but no faster.

Integrators mean problems with initial conditions, so you have to make sure the integrating capacitor starts off in the state corresponding to minimum power. The chip's soft-start facility may help with that.

Linas

Tue Jul 17 2012, 10:07AM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

hm, even if i get stable at start, it blow fuses at curie point because of fast ramping Q.

it's time for PLL
i going to change uzzros design opamps to ICL7621 since i have them around.
they should work, right ?
and also going to use 5V PLL (74HC4046) since i noticed in test that CD4046 is not 50:50 duty cycle (or i made some mistake and get that result)
74HC4046 is much closer to 50:50 than CD4046

Also can i just connect 3 to 4 leg on PLL ? since i don't use LCLR topology

uzzors2k

Tue Jul 17 2012, 05:29PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

I've also been working on and off with a new PLL driver for my series resonant induction heater. I had problems with regulation as well, and in the end I've opted to just use a shutdown comparator and reset switch. In other words the driver runs at full power until current or tank voltage exceed some limit, then BAM! the driver shuts down. It has to then be reset manually with a switch, which engages the soft-start and re-enables the driver. The phase potentiometer actually allows you to adjust phase to any desired value above or below resonance, which the driver appears to track.

1342545974 95 FT141346 Ets 4046 Ih 2 Beta

I've been forced to skip thorough testing of the driver since I need my ROV fixed and ready for next week. So if you decide to use it let me know how it works for you. If it doesn't misbehave, be a dear and make a PCB layout for it would you?

Linas

Tue Jul 17 2012, 07:00PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Uzzors wrote ...

13 leg should be 2 leg, right ?
hm, maybe i will use some ideas from your circuit :)
if some one wonders my power part looks like this:
as you can see i use LC in parallel, i don't know is it good, but voltage can rise as high as my scope can see, so > 300V

Each IGBT has RCD snubber from my old DRSSTC bridge

previous 1 2 3 4 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob